Decoding the Inverted Yield Curve

By Matt Blackman 

Since first turning negative in December, the yield curve continued to invert as 2006 began to unfold. What does that mean? Simply put, short-term yields such as the Fed funds rate, 13-week (90-day) and two-year Treasury yields moved above the ten-year to 30-year yields. Why is this important? In the past an inverted yield curve has often presaged economic slowdowns and recessions. Here’s the chain of events:

•         December 27, 2005 – the two-year moved above the ten-year yield.

•         January 31, 2006 – the Fed funds rate moved above the ten-year yield.

•         February 22, 2006 – finally, the three-month (90-day) yield moved above the ten-year as the inversion deepened.

According to one analyst, in late February  there was a 20 percent chance of recession within four quarters, given the spread between 13-week (90-day) and 2-year Treasury yields. Studies show that the longer the inversion lasts, the greater the recession risk. As March began, however, the yield curve miraculously righted itself, thanks in part to Fed language that left the door open for a pause in rate hikes. By May 1, 58 basis points separated the long end (30-year yield) from the short end (13-week yield) as long-term rates rose. (See Figure 1.) This shows the trend toward a flatter curve, the inversion that occurred in late December and early January (which is not shown in this graph since long-term rates turned up in things. First, it was an indication of market optimism that, after lifting the funds rate 15 consecutive times, the Fed was getting ready for a break, and that would be good for both markets and the economy. Second, together with the first quarter 2006 advance gross domestic product (GDP) report that showed 4.8 percent growth, it demonstrated that economic momentum was very much alive and well.

But did it mean that the risk of a slowdown had passed?

Figure 1: Yield curve: January 2004 - May 1, 2006  

Source: www.TradingEducation.com

Different This Time?

Each time an inversion occurs, economists argue that this time it’s different. They did so the last time the curve inverted in early 2000, and we all remember what followed. We heard the same comments in late 2005.  In this latest occurrence, experts discounted the risk by explaining that commodity (and food) prices have jumped due to strong demands from China and India. But it is important to take notice that average hourly earnings have fallen in real terms for the last six years. An abundance of cheap labor pools in these emerging powerhouses has put downward pressure on labor costs around the world, and unit labor costs are an important factor in inflation calculations. Because real wages have remained low, inflationary pressures are minimal; therefore, the risk of rapidly rising interest rates has remained muted.  While this comes as good news, it raises three important questions:

1. Is a recession really the event traders and investors should be striving to avoid?

2. If hourly earnings continue to experience downward pressure, won’t this ultimately have a    depressing effect on the driving force in industrial economies – consumer spending – and markets? Consumer spending represents 70 percent of the economy. The difference is that, instead of having a slowdown caused by inflation pushing interest rates higher, the greater risk is having one caused by declining consumer purchasing power.

3. If real average earnings will not be the principal driver of the economy going forward, what will take its place?

Figure 2: Yield Curve: Three-Month Minus Ten-Year Yeild

The differences between government ten-year and three-month bond yields showing inversions (red arrows). Gray boxes show bear markets in the Dow Jones Industrial Average. By the time inversions occurred, the bear market was already under way in three of seven inversions. An inversion provided advance warning of a bear market in three of seven cases as well.  The latest inversion occurred on February 22, 2006. Source: St. Louis Fed. Chart provided by www.TradingEducation.com

Best Gauge to Monitor

Let’s examine these questions one at a time. As any market watcher will tell you, by the time a recession has been confirmed, your portfolio is in tatters. This point was plainly made clear for those who read Joe Ellis’s Ahead of the Curve. Does it not make far more sense, therefore, to try and avoid bear markets when this damage is done? How accurate has a yield curve inversion been in warning of bear markets, and which inversion has provided the best advance warning? To answer this question, we looked at three yield curve relationships: three-month (13-week) minus ten-year U.S. Treasury yields, Fed funds minus ten-year Treasury yields and two-year minus ten-year Treasury yields. As Figure 2 shows, the three-month minus ten-year yield relationship inverted seven times up to 20

Figure 3: 10-Year Bond Yield –Fed Funds Rate

The ten-year minus Fed funds rate gave a better warning with a total of eight signals (red arrows) warning up to 2005 of four bear markets. Numbers show Dow Industrial percentage drops in bear markets (gray boxes) and gains in bull markets. Orange line is an eight-month moving average. A ninth inversion occurred on January 31, 2006. Source: St. Louis Fed. Chart provided by www.TradingEducation.com

In Figure 3, the Fed funds and ten-year yield curve inverted eight times between 1965 and 2005. Because data only go back to the mid-1970s in Figure 4, there are fewer examples to study. Of the five inversions that occurred up to 2005 between two-year and ten-year yields, three bear markets followed; in one case (1998) the bear market began as the inversion occurred. It is also important to note that a number of bear markets occurred without an inversion warning. In comparing Figures 2, 3 and 4, warning time varied in every chart, but you will notice that there is actually a 100-percent chance of a bear market within 26-months of an inversion in every case. Warning time was less on average, but in two cases in the three-month/ten-year and three cases with the Fed funds/ten-year chart, there were double inversions ahead of bear markets within roughly two years of the market drop. In five cases, inversions occurred after the onset of bear markets. There is also a better than 80-percent chance of a bear market within six months of an inversion.

Figure 4: 10-Year – 2 Year Bond Yields

Inversions between the ten-year and two-year bond yields provided reasonable warnings of impending bear markets between 1978 and 2005 (red arrows) in five occurrences. Since 1978, this relationship has provided the most advance warning of economic weakness. The latest signal occurred on December 27, 2005. But it is interesting to note that, like inversions between the ten-year/three month and ten-year/Fed Funds charts above, a number of times there was no inversion before a bear. Numbers show percent changes in the Dow Industrial Average through bear and bull markets. Source: St. Louis Fed. Chart: www.TradingEducation.com

Figure 5: Real Average Hourly Earnings Annualized Growth  

Monthly chart of real average hourly earnings between 1965 and early 2006 showing peaks in wage growth followed by bear markets (gray boxes). There have been 10 bear markets and six recessions (black rectangles) in that time. Also note the overall decline in real average hourly wage growth over the period. Source:  www.aheadofthecurve-thebook.com

And The Winner Is…

Which chart provided the best advance warning of an impending bear? Figure 4 appeared to provide the most reliable advance warning. In only one instance (1998) did an inversion not provide advance warning and instead occurred concurrently with the onset of a bear. The Fed funds/ten-year yield curve (Figure 3) was second, giving as many early warnings as Figure 4 but with an approximate one-month greater delay. It is important to point out that by the time a full inversion occurred (13-week/30-year), the bear market was well underway in most cases. But that does not hide the fact that, while it may take a year or more from the onset of an inversion in any of the three cases, the chances of a bear market not occurring must be considered slim for those of us who are probabilistic traders and investors.

Let’s return to out initial questions, questions two and three in particular. If real average hourly earnings, which have acted as a major driver of consumer spending and economic strength in the past, are not picking up, what will drive growth? This question cannot be answered so easily. In Ahead of the Curve, Ellis made a good case for the relationship between real wage growth (after subtracting inflation) and economic prosperity. Figure 5 shows that bear markets have often occurred following a peak in wage growth. The last major peak occurred in 1998 when wage growth topped out at an annualized rate of 3.46 percent (April). As of March 2006, real wages were growing at an anemic rate of 0.41 percent annually and had only just gone positive after 23 months in negative territory between February 2004 and January 2006. Are the dorsal-fin peaks that occurred between 2001 and 2004 warning of another bear market on the horizon? If so, it is quite late.

In Ahead of the Curve, Ellis explains what might cause a delay in economic reaction following wage earnings peaks. In the mid-1980s, a bear market (and recession) was kept at bay, thanks to the Reagan tax cuts. The same occurred in 2002-2004, thanks to the Bush Administration’s tax cuts. A 2005-06 bear market has also been repelled in large part due to the ability of consumers to borrow against rapidly increasing home prices. That has led to an extended period of prosperity, thanks to the availability of cheap money that had the effect of replacing wage growth.

Solving the Puzzle

“To dismiss the yield curve inversion out of hand by believing that this time it’s different could well prove a costly mistake,” says Darrell Jobman, senior market analyst with www.TradingEducation.com. “A far better approach is to realize that based on history, the chances of a market downturn are greatly increased, given the recent yield curve inversion and weak wage growth. The risk is compounded by an economy that has become increasingly reliant on real estate borrowing to finance consumer spending. Anything that curbs the consumer’s ability to continue spending has serious implications.”

It is also important to remember that from a technical perspective, U.S. stocks remain in a large-cap secular bear market until new all-time highs are put in and confirmed by both the S&P 500 Index and Dow Jones Industrial Average. This means that only time will tell whether the period between March 2003 and May 2006 has been nothing more than a spectacular bear market rally. Even if new highs are generated, without a solid foundation of real economic expansion driven by wages growth, any upward market momentum could well be short-lived. There is little doubt that the months and years ahead will be challenging. It is up to investors and traders to remain cognizant of these facts and act accordingly.

 Matt Blackman, market analyst for www.tradingeducation.com, is a technical trader, author, reviewer and keynote speaker.  He is a member of the Technical Securities Analysts Association (TSAA), Canadian Society of Technical Analysts (CSTA) and a  Market Technicians Association (MTA) affiliate member. He can be reached at matt@tradesystemguru.com.

Reprinted from SFO Magazine
Copyright © 2006 SFO Magazine
PO Box 849, Cedar Falls, IA 50613,  ph. 319.268.0441

Contact Matt Blackman.